Snowflake Arctic Embed M v2.0 Open-Source Multilingual Model - Supports Sentence Similarity and Feature Extraction for 50+ Languages

Snowflake Arctic Embed M V2.0

Developed by Snowflake

Snowflake Arctic Embed M v2.0 is a multilingual sentence embedding model that supports over 50 languages, focusing on sentence similarity and feature extraction tasks.

Text Embedding

Transformers

Supports Multiple LanguagesOpen Source License:Apache-2.0 #Multilingual Embedding #Sentence Similarity Calculation #Cross-language Retrieval

Downloads 92.30k

Release Time : 11/8/2024

Model Overview

This model is an efficient sentence transformer capable of converting text into high-quality embedding vectors, suitable for various natural language processing tasks such as sentence similarity calculation, information retrieval, and text classification.

Model Features

Multilingual Support

Supports over 50 languages, suitable for global application scenarios.

High-performance Sentence Embedding

Capable of generating high-quality sentence embedding vectors for various natural language processing tasks.

Extensive Evaluation Benchmarks

Performs excellently in multiple benchmarks such as MTEB, covering tasks like classification, clustering, and retrieval.

Model Capabilities

Sentence similarity calculation

Feature extraction

Text classification

Information retrieval

Text clustering

Use Cases

E-commerce

Product Review Classification

Classify product reviews on platforms like Amazon to identify positive and negative feedback.

Achieved 70.36% accuracy on the Amazon polarity classification task.

Customer Support

Duplicate Issue Detection

Identify duplicate issues in customer support platforms to improve efficiency.

Achieved 60.65% MAP on the AskUbuntu duplicate question task.

Academic Research

Paper Clustering

Cluster academic papers to help researchers discover related studies.

Achieved 44.68% V-measure on the paper clustering P2P task.

🚀 snowflake-arctic-embed-m-v2.0

This model is designed for sentence - similarity tasks. It supports a wide range of languages and has been tested on multiple datasets from the MTEB benchmark, providing various evaluation metrics.

📚 Documentation

Model Information

Property	Details
Pipeline Tag	sentence - similarity
Tags	sentence - transformers, feature - extraction, sentence - similarity, mteb, arctic, snowflake - arctic - embed, transformers.js
Languages Supported	af, ar, az, be, bg, bn, ca, ceb, cs, cy, da, de, el, en, es, et, eu, fa, fi, fr, gl, gu, he, hi, hr, ht, hu, hy, id, is, it, ja, jv, ka, kk, km, kn, ko, ky, lo, lt, lv, mk, ml, mn, mr, ms, my, ne, nl, pa, pl, pt, qu, ro, ru, si, sk, sl, so, sq, sr, sv, sw, ta, te, th, tl, tr, uk, ur, vi, yo, zh
License	apache - 2.0

Model Results

snowflake - arctic - embed - m - v2.0

Dataset	Config	Task Type	Main Score	Other Metrics
MTEB AmazonCounterfactualClassification (en - ext)	en - ext	Classification	66.6867	accuracy: 66.6867, f1: 55.0373, f1_weighted: 73.07430000000001, ap: 18.077399999999997, ap_weighted: 18.077399999999997
MTEB AmazonCounterfactualClassification (en)	en	Classification	66.194	accuracy: 66.194, f1: 60.854299999999995, f1_weighted: 69.57339999999999, ap: 30.279099999999996, ap_weighted: 30.279099999999996
MTEB AmazonPolarityClassification (default)	default	Classification	70.3589	accuracy: 70.3589, f1: 70.0409, f1_weighted: 70.0409, ap: 64.81949999999999, ap_weighted: 64.81949999999999
MTEB AmazonReviewsClassification (en)	en	Classification	33.766	accuracy: 33.766, f1: 33.3656, f1_weighted: 33.3656
MTEB ArguAna (default)	default	Retrieval	58.011	ndcg_at_1: 33.144, ndcg_at_3: 47.909, ndcg_at_5: 52.932, ... (a long list of metrics)
MTEB ArxivClusteringP2P (default)	default	Clustering	44.684400000000004	v_measure: 44.684400000000004, v_measure_std: 13.5064
MTEB ArxivClusteringS2S (default)	default	Clustering	35.0503	v_measure: 35.0503, v_measure_std: 13.9543
MTEB AskUbuntuDupQuestions (default)	default	Reranking	60.648500000000006	map: 60.648500000000006, mrr: 74.528, nAUC_map_max: 19.4239, nAUC_map_std: 20.0729, ...
MTEB BIOSSES (default)	default	STS	87.0568	pearson: 89.5081, spearman: 87.0568, cosine_pearson: 89.5081, ...
MTEB Banking77Classification (default)	default	Classification	80.1818000	accuracy: 80.1818000, f1: 79.5538, f1_weighted: 79.5538

📄 License

This project is licensed under the apache - 2.0 license.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご